Complementary classification of hypertrophic cardiomyopathy using unsupervised cluster analysis on left ventricular function

Contemporary classification of hypertrophic cardiomyopathy (HCM) was mainly based on the site of myocardial hypertrophy and left ventricular outflow tract obstruction. A complementary classification based on left ventricular function could provide a powerful tool to identify individuals with high risk of adverse cardiovascular outcomes and guide individualized managements. Multi-dimensional echocardiographic parameters of left ventricular function derived from conventional echocardiography, tissue Doppler imaging, and speckle tracking echocardiography were obtained in 266 HCM patients and 169 healthy controls (HCs). According to these parameters, HCM subtypes were calculated by principal component analysis and unsupervised cluster analysis. Variables of different groups were compared. The prognosis between HCM subtypes were evaluated. There were two HCM subtypes generated, subtype 1 HCMs (n = 123) and subtype 2 HCMs (n = 143). Compared to HCs, left ventricular diastolic and systolic function were significantly declined to varying degrees in both subtype 1 HCMs and subtype 2 HCMs, especially in subtype 1 HCMs (all P value < 0.001). Subtype 1 HCMs characterized as increased LAVI and E/E', decreased mean E' and untwist rate, increased global and segmental longitudinal strains, circumferential strains and radial strains, decreased rotation degree, twist degree, and twist rate, in comparison with subtype 2 HCMs (all P value < 0.001). Notably, subtype 1 HCMs were more susceptible to adverse prognosis of atrial fibrillation (HR: 4.34; 95% CI 1.08-17.53; P value: 0.039). Collectively, we stratified HCM patients into two subtypes with different diastolic and systolic performance and risk of atrial fibrillation. This complementary classification might eventually help to target management of HCM patients who would benefit most.